A sourcing agent misclassified one of us by 3,300 miles. Today, we are building security infrastructure for high-assurance AI to prevent mistakes like this at industrial scale.
CEO
Strahinja Janjušević
CEOStrajo came from Montenegro to study Offensive Cyber at the US Naval Academy, before conducting research at NASA, Microsoft and VectraAI. At MIT, he researched AI-enabled cyberattacks and ways to secure maritime cyber-physical systems with AI.
COO
Adam Flanagan
COOFormer AI-focused VC at Creator Fund, leading Applied AI and cyber investments across Europe. Adam built production LLM agents for deal sourcing and portfolio ops, which is when he saw the failure modes Refractal now exists to solve.
AI Research Engineer
Ruchira Dhar
AI Research EngineerPhD researcher in AI at the University of Copenhagen and visiting researcher at ETH Zurich. Builds the science of evaluating intelligent systems, with a focus on generalisable and efficient evaluation. Prior to academia, Ruchira shipped production-grade AI in industry and built test suites for real world model deployments.
Founding Research Engineer
Abdelrahman Hekal
Founding Research EngineerAbdel specialises in AI safety and formal verification. He completed his PhD at Newcastle University and previously worked on verifying neuro-symbolic AI at Imperial College London’s Safe AI Lab.
Founding Engineer
Wee Joe Tan
Founding EngineerJoe is a Computer Science student at UCL who previously managed cloud infrastructure in the Singapore Army. His full-stack software engineering experience includes projects at GovTech, OpenSRE and CSIT.
Forward Deployed Engineer
Djordje Ivanovic
Forward Deployed EngineerHarvard computer science student and AI engineer working across agent systems, government automation, and secure AI infrastructure. Djordje has built production AI tools, national-scale data systems, and technical partnerships, bringing a product and deployment perspective to Refractal’s work.
Managing Director · Accenture
Ysais Martinez
AdvisorLeads Accenture’s global Microsoft Security business, scaling MxDR and SecOps modernization across Defender, Sentinel, and Security Copilot for enterprise clients. CISSP & CCSP; deep operator background across identity, cloud, and managed detection.
CISO · US Military Academy
Jonathan Fuller
AdvisorCISO at the United States Military Academy at West Point and Assistant Professor, with a PhD in cybersecurity. Focuses on adversary behavior, detection engineering, and AI security, with a background in secure communications and enterprise systems.
Sr. Security Lead · Microsoft
Andrew Shea
AdvisorSenior Security Lead at Microsoft, partnering with US Government and DoD customers across identity, cloud, and SecOps. US Naval Academy graduate; founder of Operation MOS, an AI-powered resume builder for transitioning veterans.
Google Cloud Global Architect
Andrew Liebowitz
AdvisorGoogle Cloud Global Architect at Google, helping enterprises roll out Workspace with the right identity, device, and data controls. SANS Technology Institute graduate carrying GIAC certifications across incident handling, security essentials, and foundational cyber.
Harvard LLM - AI Policy and Privacy Expert
Tea Mustać
AdvisorCOO of SafePorter and an AI governance expert advising startups and organizations on responsible AI strategy, agentic systems, data protection, and technology regulation. A Harvard Law School graduate, she helps teams design AI systems that are compliant, commercially viable, and human-centered. Co-author of The AI Act Compact and co-host of RegInt: Decoding AI Regulation.
Harvard LLM - EU Cyber, AI & Privacy Law
Henrik Nolte
AdvisorDoctoral candidate at the University of Tübingen and graduate of Harvard Law School, specializing in EU cybersecurity, AI, privacy, and biotechnology law.
VectraAI · Data Scientist
Anna Baron Garcia
AdvisorSenior Data Scientist at Vectra AI with a Ph.D. in Electrical Engineering & Computer Science (2022) focused on machine-learning models for cybersecurity engineering. Her current research centers on AI agentic cybersecurity engineering for the network security ecosystem (AI red teaming).
Adam was an investor at Creator Fund, backing early-stage Applied AI and cyber founders across Europe. Alongside the day job he was building his own LLM agents for sourcing founders and scaling his own work. He spent more time patching the agents’ mistakes than using their output.
Strajo was on the other side of the Atlantic. After coming from Montenegro to study Offensive Cyber at the US Naval Academy, he joined MIT to research Adversarial Intelligence. He spent time in research across Microsoft, NASA, and VectraAI, where he wrote a paper using Claude for adversarial red-teaming.
Adam’s sourcing agent detected Strajo’s paper and flagged him as a potential founder, in Cambridge, UK. Except he was in Cambridge, Massachusetts.
Adam only spotted the mistake when manually reviewing the Excel output. We started Refractal to stop these subtle failures that happen today on an industrial scale.
What the sourcing agent returned. Confidently. With a Zoom invite.
Where Strajo actually was. 3,314 miles off. No policy check stopped it.
Refractal operates from London, a city that has been regulating risk for three centuries.
